I am studying diversity in the nosZ gene in marine bacterial communities. Sequencing was done using Illumina Miseq, and after some preliminary analysis (e.g. quality trimming using Pear) I now have a large fasta file containing over 500,000 nucleotide sequences. Before I move on with the analysis, I want to make sure that all sequences in this file are actually nosZ gene sequences. However, I am not sure how to proceed.
Any suggestions would be really appreciated.
I can think of a few ways to do this.
1) The simplest is to blast your genes against a database of known NosZ genes and only take genes above a certain cutoff.
2) The second is to use HMMER to run a NosZ-specific HMM over your dataset and count the good hits (see fungenes for a pre-built NosZ HMM http://fungene.cme.msu.edu/).
3) A more interesting way to go about this is MetAnnotate (http://www.biomedcentral.com/1741-7007/13/92), which will place your sequences into a phylogenetic tree built on NosZ genes and give you an annotation and phylogenetic placement. This is the only way I can think of to see whether your NosZ genes fall into a group outside of the known functional NosZs...i.e. maybe they're just closely related sequences that aren't really NosZ, or maybe they're a new clade of NosZ.
- Badges
- Science method